When news about the $700 billion bailout plan first came out last month, the U.S. stock market staged a big rally on September 18 and 19. The S&P 500 index rose 8.5% in two days. Respected author Larry Swedroe posted this on the Bogleheads forum on Sunday, September 21:
“It is often discussed about TLH  whether to wait 31 days and reinvest or immediately.
“As you know I have been proponent of not waiting for two reasons–every day you are out of market you miss the ERP . But you also have far greater chance of missing a big up then avoiding a big down. Using 5% as the definition of “big” I found 70% more likely to miss the big up than avoid the big down
“Imagine harvesting losses a few days ago and sitting it out now. Already missed about 8% –of course we dont know the eventual outcome.”
 TLH: Tax Loss Harvesting. The practice of selling an asset for a tax loss and then either immediately purchase a similar asset or wait 30 days before buying back the same asset.
 ERP: Equity Risk Premium. Stocks have a higher expected return than cash or bonds.
Yes, imagine that. I bet whoever sold on Sept. 17 for tax loss harvesting and decided to sit out in cash for the next 30 days had NO idea what kind of a dizzying fun ride he was going to miss.
Today is Oct. 20. Our hypothetical investor who sold for tax loss on Sept. 17 can buy back the same fund now without triggering a wash sale. The trading day just started but it looks like our investor can buy back at a price at least 15% lower than what he sold at. This once again shows how unreal the argument is for missing the best days in the market. Our investor missed the best day in the market in 70 years. His investment return not only didn’t suffer but it will actually be higher.
Larry Swedroe argued that because there is an equity risk premium, for each day an investor is out of the market, he loses the equity risk premium. Well it doesn’t work neatly like that. While there is an equity risk premium, the stock market does not go up in a straight line. In a short period like 31 days, the chance of the stock market going down is as good as the chance of it going up, especially when the market is already down which triggered the tax loss in the first place.
I pulled the daily closing prices of Vanguard Total Stock Market ETF (VTI) in 2008 and compared them with prices 31 days earlier. If an investor sold for tax losses and sat in cash for 31 days, 54% of the time this investor would see a lower price on day 31. For the other 46% of the time when the price was higher on day 31, the investor didn’t have to wait too long before he would see a lower price again, at which point he could complete the repurchase for tax loss harvesting and suffer no loss from being out of the market or even profit from it.
|Additional Wait after 31 Days for a Lower Price||Cumulative %|
|up to 10 days||64%|
|up to 30 days||78%|
|up to 60 days||91%|
|up to 90 days||100%|
If you’d like to look at more data from this unscientific sample, the spreadsheet is here:
Spreadsheet: Tax Loss Harvesting: Wait 31 Days or Buy Similar Fund
What if someone sold on October 10, right before the 11.6% jump on October 13? Let’s see how long it will take for a lower price to show up this time. I’m afraid it won’t be too long.
Say No To Management Fees
If you are paying an advisor a percentage of your assets, you are paying 5-10x too much. Learn how to find an independent advisor, pay for advice, and only the advice.
If your goal is to harvest losses without affecting your asset allocation then you should immediately repurchase a similar asset.
If your goal is to time the market using a “sell low, buy lower” strategy, then follow whatever strategy you like. You may want to base your conclusions on more than hindsight in the current bear market, though.
As TSM is still on a downward trend this year, you data is skewed towards holding the money in cash for 31 days. Your conclusion would probably be different for 2003, 2004, 2005, 2006, 2007 or if TSM starts recovering for the rest of 2008.
The fact is nobody knows when the bottom will be reached, and for me, I find it hard to go to cash for the 31 days.
Harry Sit says
Jim, IndexFundFan – The purpose of this post is not to show that waiting 31 days in cash is a superior strategy or how to time the market. Rather it’s meant to show that buying a similar fund immediately is not necessarily better. Larry was trying to make a point with two days of hindsight — “see these people missed the best days.” The actual outcome proved him wrong, big time.
Stocks go up in the long term. We can have some degree of certainty of saying stocks will be up in 30 years. But our degree of certainty comes to close to a coin toss if we are talking about a short period like 31 days. Just for kicks, I’m going to pull the same data for 2003-2007 and see what it looks like in years when the market went up. Stay tuned. Keep in mind though tax loss harvesting opportunities don’t come up as often in bull markets.
You put out lots of good work and I appreciate your site, but I disagree with the bulk of this post.
I would think that denying that the investor loses out on the ERP would be equivalent to either believing that your evaluation of market value is better than the markets OR that stocks will not have a positive return going forward. What stock market theory would suggest that your assumption is true?
Think about Larry’s ERP statement in an expected value sense. If you TLH every year, you are only in the market for 11/12 of the time, why would you think you would receive anything other than 11/12 of the ERP.
And 12:11 is not a coinflip.
Of course Larry’s statement looks bad over the past 10 months, VTI is down 30%! As others have said, I would recommend gathering as much data as possible, you can get 20 years of vfinx on yahoo, probably do even better if you are more resourceful. The calculation is just a few click and drags whether you calc for 1 year or 100.
Lastly, depending on your wash sale rule interpretations and the price you pay for trades, buying a similar fund without waiting can be very low risk. I was hoping this post might address alternatives to avoid any wash-sale worries. For example, you wouldn’t replace VFINX with FSMKX, but you might be comfortable replacing it with VTI, VV, or MGC. Suggesting best-alternatives for TLH could be an interesting post.
Keep up the good work. I appreciate it very much even if I only comment when I disagree.
Harry Sit says
Matt – No problem with only commenting when you disagree, and thank you for cutting me some slack. However I didn’t say either of the things you said I said. I didn’t say I have a better judgment of value than the market, nor that stocks don’t have a positive return. I only said in the short term, it’s unpredictable. Nobody can predict what the stock market will do in the short term. Even Warren Buffett said “I have no idea what the stock market is going to do next month or six months from now.”
Let me state the problem in a different way. Why is someone afraid of staying in cash for 31 days? Because he/she is afraid that the bear market bottom is going to appear in the next 31 days and the market will be higher forever after 31 days are over. It’s this fear of missing the boat, fear of missing the best days. If that’s not the case, this investor can always get back to the same fund at the same or lower price point after 31 days. Now, what do you think the chance is for any investor to successfully predict that the bear market bottom will appear in the next 31 days? Not good, precisely because nobody can call the bottom.
Thanks for the response. I may have communicated my points in a poor way,
“I would think that denying that the investor loses out on the ERP would be equivalent to either believing that your evaluation of market value is better than the markets OR that stocks will not have a positive return going forward. ”
Here I was suggesting that the only way missing 31 days could be a coin-flip would be if one of the quoted assumptions were true. I was not trying to imply you said these things.
“What stock market theory would suggest that your assumption is true?”
I should have clarified, by “your assumption” I meant the assumption that missing 31 days was a coin flip.
I actually regret even including that first paragraph in the post, I think it is a more convoluted way to make my point, which is:
If the stock market goes up over time, and you cannot time the market, any period will tend toward a positive expected return.
And I think that is taking Buffett out of context. Let’s pretend the IRS changes the rule to 6 months. Would you be comfortable using the same logic to stay out of the market for 6 months? It must also be a coin flip?
To answer your questions, I would be afraid of staying in cash because I would lose 1/12 of the ERP and I perceive very little risk to TLH with similar but different funds. I wouldn’t be concerned about timing the bottom, because I do not try and time the market. If I TLH and the market went 10% lower the next week I would consider the risk-reward tradeoff of TLH once again. I don’t think predicting the bottom really has much to do with this discussion either. Since we cannot time the market, TLH should always be evaluated at current prices, if the market goes lower, re-evaluate.
Harry Sit says
Matt – You asked “Let’s pretend the IRS changes the rule to 6 months. Would you be comfortable using the same logic to stay out of the market for 6 months? It must also be a coin flip?” I think the time gap definitely makes a difference. If the rule is changed to 6 months, it will tilt the balance to buying a similar fund. If it is changed to 6 years, it will tilt even more. On the other hand, if the wash sale rule is changed to 1 week, 1 day, 1 hour or 1 minute, do you still bother buying a similar fund? I don’t know precisely where the time cutoff should be. It’s not going to be black-and-white anyway. One month feels short to me. You probably think it’s quite long. But between 1 minute and 31 days, you must also have a cutoff somewhere, no? So I guess it comes down to our comfort level on how long is long and how short is short.